Dynamic Self-Triggered Parallel Control of Guidance Intercept Systems With Constrained Inputs via Adaptive Dynamic Programming

This paper introduces an anti-saturation optimization algorithm based on dynamic self-triggered adaptive dynamic programming (ADP) for bounded acceleration guidance interception. By establishing a nonlinear input-constrained guidance intercept and control system, the smooth bounded function is used to constrain the system input to ensure that the system operates in a controllable range. Subsequently, the appropriate performance function that can accurately reflect the system is designed. In alignment with the advantages of parallel control and ADP, a group of parallel systems are constructed by modeling the derivation of control input. A self-learning control framework is explored, facilitating virtual-actual interaction and mutual reinforcement between multiple controllers, optimizing the management of the interception system. Furthermore, event-triggered control (ETC) policies are devised, which can be updated only when needed by setting trigger conditions, thus saving data resources. The stability proof of the closed-loop system is given to ensure that the system can keep stable operation when the trigger condition is satisfied. On this basis, a dynamic self-triggered control (DSTC) with soft computing is put forward, enabling trigger instant calculation without real-time system monitoring and further extending the trigger interval. Simulation results evidence that the devised guidance scheme can intercept the maneuvering target at a reduced expenditure of communication resources.